The present invention relates to a method for checking consistency of distributed data by detecting and eliminating inconsistent data in a system in which a plurality of devices connected to a common data transmission line process data cooperatively.
In Japanese Patent Unexamined Publication No. 57-97173, a plurality of devices connected to a common data transmission line sequentially edit and add distributed files by connecting files which the devices for a request to edit the distributed files. In this system, if a file is stored in duplicate in a plurality of devices in order to improve reliability of the file edition, redundant edit data is generated for a request for file edition. When the devices sequentially edit the data, each device selects correct data from the redundant edit data to add the selected data to the file of the device so that the reliability of the distributed file edition is improved.
On the other hand, in Japanese Patent Unexamined Publication No. 57-175238, a method to select correct data from redundant data is disclosed. In this method, a device which receives the redundant data collects data for a predetermined time period, compares the data contents and selects the correct data by majority logic. By applying this method to the distributed file edition system described above, the reliability of the distributed file edition is improved by storing the files in duplicate in the plurality of devices.
However, if such distributed files are sequentially edited, the size of the edited data sequentially increases. Accordingly, when the device selects the correct data from the redundant edit data by comparing the data contents as disclosed in the Japanese Unexamined Publication No. 57-175238, the data to be compared is larger, the compare time is longer and the load is heavier.